Method of Investigation. 17 







ment, are at present the best that can be obtained, although it is well recog- 

 nized by physiologists that the determination of the heat of combustion of 

 urine is one of the most perplexing problems with which we have to deal. 



In preparing samples of food and feces for combustion, it is necessary first 

 to partially dry the substance, as otherwise the combustion would be very un- 

 satisfactory. In this partial drying, loss of organic matter, and consequent 

 loss of potential energy, should be avoided. Our experiments with vacuum 

 desiccators have again shown the wisdom of drying physiological preparations 

 and food materials at room temperature in a vacuum rather than at a tempera- 

 ture much above that of the body, i. e., the water-bath or steam-oven. 



Heat elimination from the tody. The chamber of the respiration apparatus 

 is provided with appliances for measuring heat. These appliances have been 

 described in detail in a previous publication. 48 The respiration chamber, which 

 consists essentially of an air-tight copper box, is surrounded with alternate 

 layers of air, zinc, and wood, so as to minimize the radiation. The heat pro- 

 duced by the subject is brought away by a current of cold water flowing through 

 a copper pipe, to which a large number of disks are soldered to increase the 

 absorbing surface. Arrangements are made, by heating and cooling the air 

 surrounding the chamber, to render the walls of the chamber adiabatic. By 

 noting the rise of the temperature of the water in passing through the cham- 

 ber, and the mass of the water, the amount of heat brought away by the water- 

 current can readily be computed. The heat required to vaporize the water 

 given off in the air-current is obtained by multiplying the number of grams 

 of water vaporized by the factor 0.592. 



Complete tests of the respiration calorimeter. Burning known weights of 

 ethyl alcohol in the chamber furnishes a means of checking the accuracy of 

 the respiration calorimeter for measuring water and carbon dioxide output and 

 oxygen intake, as well as heat elimination. A large number of such check tests 

 have given very satisfactory results. In general the quantities actually deter- 

 mined differ from the calculated amounts by less than 1 per cent. 



RECORDING RESULTS AND THE USE OF DECIMALS. 



In the numerous computations involved in experiments of this nature it is 

 frequently difficult to determine early in the calculation just what degree of 

 refinement in the mathematical calculations is warranted by the accuracy of 

 the experimental process from which the data are derived. Such decision has 

 been withheld in every instance till as late as seemed necessary, in some in- 

 stances undoubtedly too late, and hence the results are at times expressed with 

 one or two decimal places that are not of real significance. It is to be noted, 

 however, that in the final summary and balance tables supernumerary figures 

 are in general omitted. 



48 Carnegie Institution of Washington Publication No. 42 (1905). 

 2 



